cutimage - Fotolia
What to expect from application performance monitoring tools
Performance monitoring tools preemptively thwart run-the-business application problems, but only when you have the right APM tool for the job.
Users typically don't recognize the value of run-the-business applications until there's a problem.
Without effective performance management, applications suffer response time delays, anger customers and ruin employee productivity. Trial-and-error troubleshooting and poor visibility into problems cause outages. Both situations make key business functions effectively unavailable for long periods of time, choking off the sales and production cycle.
In-house performance management tools struggle to keep up with run-the-business applications. Complex apps, such as enterprise resource planning, may add tiers, new software types such as cloud computing interfaces and features to maintain performance as storage sizes grow by more than 50% per year. These changes make application performance monitoring and management (APM) much more difficult.
Some IT shops, or the businesses they serve, underestimate the need for effective APM, or fail to anticipate the effects of increasing complexity on application performance.
High-performance tools
Sophisticated application performance monitoring tools cover all aspects of the app's distributed architecture, and adapt flexibly to new technologies and new additions. A cost focus embedded in the tool could guide application architecture redesigns, minimizing long-term costs while eliminating performance inhibitors.
An APM tool must discover root causes and then identify fixes rapidly. Look for tools that can troubleshoot proactively as well as reactively. Are problem reports easy to use and on-point? Can you drill down deeply into any tier of the architecture, or in the network, at multiple levels of the software stack (application, application server and database)?
Monitoring tools proactively prevent problems by generating a historical database of performance trends, allowing problem isolation to a single instance of a particular set of software. For example, Idera's Precise Software for SAP includes what-if tools that show how an infrastructure change, such as adding solid state drives, will alter application performance.
Don't let the performance monitoring and management tool be what's degrading the application's speed in the first place. Choose tools with low overhead on the system. A lightweight (e.g., log-based) architecture minimizes overhead while maximizing collected data.
The performance payoff
Good application performance monitoring tools cause surprising jumps in customer and business/end-user satisfaction. IT teams can deliver up to an order-of-magnitude faster problem correction and cut out trial-and-error outages. Relatively small glitches are caught and rectified before they become big problems.
The tools help teams prevent problems, not just solve them. Through performance trends analysis, IT teams can lessen the burden of architecture upgrades and make more effective and cost-effective use of hardware. Without performance management, expect to see IT always fighting fires and applications suffer at key times.
About the author:
Wayne Kernochan is president of Infostructure Associates, an affiliate of Valley View Ventures, with more than 20 years of experience in IT industry analysis. This document is the result of Infostructure Associates-sponsored research. Infostructure Associates believes that its findings are objective and represent the best analysis available at the time of publication.
Your first task, I'd think, would be to analyze the app problems so that you can work out the best solution.
If it can be fixed with a code rewrite, or attacked with a code fix, that's probably the quicker and cheaper solution. If it can't, you may be able to get around the problem with more robust hardware.
Consider both solutions. Repair the surface problem with a rewrite. Keep your enterprise safer behind stronger storm doors.